arm: Remove bogus comment in futex_atomic_cmpxchg_inatomic()
commit 522d7dec(futex: Remove redundant pagefault_disable in futex_atomic_cmpxchg_inatomic()) added a bogus comment. /* Note that preemption is disabled by futex_atomic_cmpxchg_inatomic * call sites. */ Bogus in two aspects: 1) pagefault_disable != preempt_disable even if the mechanism we use is the same 2) we have a call site which deliberately does not disable pagefaults as it wants the possible fault to be handled - though that has been changed for consistency reasons now. Sigh. I really should have seen that when committing the above. :( Catched-by-and-rightfully-ranted-at-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> LKML-Reference: <alpine.LFD.2.00.1103141126590.2787@localhost6.localdomain6> Cc: Michel Lespinasse <walken@google.com> Cc: Darren Hart <darren@dvhart.com>
This commit is contained in:
parent
6e0aa9f8a8
commit
07d5ecae29
|
@ -97,9 +97,6 @@ futex_atomic_cmpxchg_inatomic(u32 *uval, u32 __user *uaddr,
|
|||
if (!access_ok(VERIFY_WRITE, uaddr, sizeof(u32)))
|
||||
return -EFAULT;
|
||||
|
||||
/* Note that preemption is disabled by futex_atomic_cmpxchg_inatomic
|
||||
* call sites. */
|
||||
|
||||
__asm__ __volatile__("@futex_atomic_cmpxchg_inatomic\n"
|
||||
"1: " T(ldr) " %1, [%4]\n"
|
||||
" teq %1, %2\n"
|
||||
|
|
Loading…
Reference in New Issue